ElectroMagnetic Analysis and Injection of Secure Circuits – EMAISeCI

The security constitutes a crucial component of Media and Communication technologies. It represents, without any doubt, one of the levers of its rise because it is at the base of the introduction of confidence necessary for the end users. Among the threats which weigh more on the security, the vulnerability of the electronic material which implements cryptography in particular to return the services necessary of confidentiality, identification and authentification is perhaps most important. Indeed, certain fraudulent handling or attacks on this hardware makes it possible to extract from the confidential information like the keys of encryption and thus to put at evil all the protected transmission chain of information. The track race engaged between the designers of circuits and the badly disposed people accelerates with the diversity of the systems, their opening and their multiplicity. A major stake in the security of the communication systems is to improve resistance of the components to these attacks. If the development scheme, based on a very pragmatic approach of the problem, set up for the first products establishing of the functions of security were an acceptable time, it is not it today any more.
Among the most known attacks, those called by "side channels" (or observation), exploit the correlation between the handled data and the consumption or the electromagnetic radiation of the component. Another type of attacks, called by "injection of faults" circumvents protections intended to protect the sensitive informations, while modifying, for example by illumination or glitching the supply voltage, the operation of the component. A third type of attacks, more difficult to implement, consists in analyzing the design of the chip using invasive processes (abrasion, chemical engraving, laser, SEM, etc) then to probe the signals carrying secret information. It thus seems today a major stake in the security of the communication systems, to improve hardness of the components to these attacks. This stake is even more important for embedded technologies. Indeed, once this type of product on the ground, its cycle of "survival" is responsibility for the seller. He must be able to estimate, by advance, the potential resistance of its systems to the vulnerabilities of tomorrow. The degree of robustness of a circuit is thus, more than ever, a parameter determining for the manufacturers of circuits, it is one to put forward to them and an important witness of their competitiveness.
If the electromagnetic susceptibility of the circuits is largely studied by CEM community, no work is dedicated to the exploitation of this one intentionally to generate faults within the protected circuits and to carry out differential analyses of faults (DFA). The field of investigation “electromagnetic emissions and security” thus seems virgin. Moreover, in order to reduce consumption and to make possible the thinning thicknesses of oxide, the decrease in the supply voltages led to a higher vulnerability of the components. With each technological jump, the components become sensitive to increasingly fast transitory signals for given amplitude. The objective of the EMAISeCi project is to allow a theoretical comprehension of the influence of EM (for the observation or the injection of faults) on integrated circuits; in that, it is distinguished clearly from the work devoted to the study of the electromagnetic compatibility of electronic devices. This comprehension will later on make it possible to build counter measures dedicated to the emergent threats on security based on the exploitation of the EM channel.
This research project is a fundamental one. It will make it possible to confront the scientific assumptions put forth with the experimental results obtained thanks to the developments of advanced security characterization benches.